It has been proposed that erythrocytes, infected by one male and one female gametocyte, enhance malaria transmission by lowering encounter time between male and female gametes once inside the mosquito vector. This may have important implications if they occur in human Plasmodium infections. Double gametocyte infections (DGIs) have been found in Plasmodium cultures, but it is thought that they are an artefact due to the artificially high crowding of cultures. Here, we studied gametocyte density and DGI occurrence in Haemoproteus columbae infecting feral pigeons (Columba livia), to determine if crowding is the key factor producing DGIs. We demonstrate that DGIs are not a spurious phenomenon or an artefact of crowding, but occur in any gametocyte density in a proportion a bit higher than that expected by a Poisson distribution. © Springer-Verlag 2005.